Dec 18, 2007

Some thoughts on FLL competition... Part II

Some of the comments/discussions going on in the previous post got me thinking a little more and, of course, I realized after posting it that there were some things missing. So, apologies in advance for another long post, here goes:

1. Table construction

The table I was a referee for (2 tables, actually) were not matched properly. One table was about 1/8 to 1/4" lower than the other and this caused the satellite to be angled lower on one side and higher on the other. There were quite a few teams that "missed" the trigger because there robot was designed for a perfectly horizontal surface and the trigger at a very specific height from the table's bottom. Obviously, this mismatch in table heights caused some headache, so I gave many teams the benefit of the doubt and awarded the points (and the satellite pointer) if I was able to determine they would have triggered it if the satellite had been at the correct angle (equal on both sides). The point here - inspect your table before you begin and let the referee know if you have any concerns.

2. Trays on table

We (all refs) did not allow trays to be set in the parking lot. This caused some controversy as any team that wished to use some sort of tray or box to have to hold it or sit it on the floor. Most teams just placed the correct pieces in the parking lot, but for some reason this ruling caused some irritation. If this is a key point to your team's success, discuss it with the refs before competition starts and find out the ruling. (As I understand it, trays are allowed but non-LEGO parts are NOT allowed to be set on the table - before everyone rushes in on this one, let's keep the comment to 1 or 2 AND only if you have an official answer and can provide a reference)

3. Sleep Timer

I had one or two robots turn themselves off during the waiting period before a match - this obviously caused the team problem when the match started because they thought the robot was turned on and ready to go. REMEMBER: either set your sleep timer to 30 minutes or 1 hour or off completely - you don't want to risk your robot turning off while you wait and then having to go through the power up while the clock is ticking.

4. Rescue

Find out what your referee's position is on robot rescue. If the referee or a team member touches the robot outside of base, you lose a barrel as a penalty. Find out if the referee will allow the team member to reach for the robot or if the ref prefers to do it. Most refs will allow the team member to reach for their own robot if they say they are doing so - otherwise a request to the ref for a rescue might be required. Again, ask and know the procedure.

5. Sensors... again

Yes, I forgot that every NXT team is using the rotation sensor built into a motor. But when I was talking sensors, I was specifically referring to either the Ultrasonic or Light or Touch. I've said it before and I'll say it again, the Touch sensor makes a great "start" button and can replace the Orange Enter button on the front of the Brick. I saw MANY MANY teams reaching around to press the Orange button and not get their hands pulled away fast enough causing the robot to bump off course and forcing the team to grab and restart. If you've got a spare Touch sensor, mount it on top or to the side of your robot. In your program, you simply place a WAIT-TOUCH block (Orange block with the Hourglass and Touch Sensor symbols on it) and program it to wait until the Touch sensor is bumped (pressed and released)... then any remaining programming blocks in the program will begin to execute. This gives you time to pull your hand away without interfering with your robot's movement AND keeps you from having to reach around in an awkward position to push the Orange button.

6. Gears

Many teams asked about gears - they knew that gearing could give a robot more speed using certain gearing (and more power using other gearing options). This isn't the place to provide that information, but rest assured that there are plenty of places on the Internet that can show you how to use gearing with your NXT robot to give you either more speed or more pushing power.

7. Professional behavior

I cannot praise the teams I saw enough regarding their behavior at the competition. Teams were happy to be there and I saw ZERO trash-talking and/or negative comments from any teams. I think this was what made the event so successful - all the adults I spoke with that day were just smiling and laughing at how much fun the kids were having. I know it's a contest, but there was a team there that was high-fiving one another after completing ONLY the two missions they set out to complete. They weren't there to win but to show their parents/coach and everyone else that they were successful in making their robot do something over and over again. They chose to ignore the contest and focus on very specific obstacles that they decided to try and tackle. In the process, they learned some very important building and programming concepts and can apply those to future tasks.

8. Ingenuity

I think my earlier post may have misled people into thinking there was no innovation in terms of robot building. Completely untrue. Whereas sensors were missing from 95% of the robots, there were some true engineering marvels seen that day in terms of construction and execution. I saw some solutions to certain missions that I honestly NEVER would have thought of but now I have some new tricks in my bag after observing them. It never ceases to amaze me how a team can come up with a solution that no one has ever considered. I would estimate that out of the 30 robots I saw that day, about 5-10 of them had an innovative mechanism or method for solving a mission.

9. Combination Missions

Someone emailed and asked about teams performing two or more missions at a time. Yes, this did happen, but not that often. When I saw it, it was mostly the robot starting out by doing the satellite mission and then turning left or right to do one additional mission. I did see 1 team during the entire day that attempted (successfully) 3 missions in one shot (won't tell you which ones to protect that team's strategy). Given the time limit on the missions, I do think that the winning teams at International this year are going to absolutely have to combine missions in order to reduce time at base. Time in base may be only 5 or 10 seconds, but do that 5 or 6 times and you've lost 1 minute of competition time. Think about it.

10. Attachments

I believe that 90% of the teams I saw used some sort of attachment in one or more missions. By attachment, I'm talking about something easily and quickly removed at base in order to prep the robot for the next mission. There were a couple of robots that were completely stand-alone and required no additional work (they were larger in size, obviously). I even saw a few Tribots in the mix! Some teams get their kits late in the game and there is no reason to reinvent the wheel; if the Tribot works for you as a good base, go for it. It certainly is capable of holding and using attachments as the NXT-G software exercises demonstrate. Either way, the variety of attachments I saw was large - I saw cages, pincers, launchers, lances, etc. Keep in mind, though, that an attachment is only as good as it is structurally sound. If it's flimsy and comes apart easy in your hands, imagine what the satellite or house are going to do to it if they collide.

Jim

13 comments:

Joshua Heinzl said...

1. Table construction

1/8" to 1/4" in height different is pretty large! This would not have affected us, but still wow! I think the table instructions should be followed a little more carefully:
http://www.firstlegoleague.org/default.aspx?pid=29580

3. Sleep Timer

There are ways to fix this other than changing the default sleep timer.

4. Rescue

Team members are allowed to rescue the robot. See rule 21:
http://www.firstlegoleague.org/default.aspx?pid=29610

5. Sensors... again

The built-in motor encoders can be used for more than just odometry navigation! That's all I can comment on this one though ;)

6. Gears

The NXT does not encourage gears in many situations unfortunately.

7. Professional behavior

Professional behavior this year is, in my experience, much better than in previous years. Most teams are congratulatory and well spirited. This certainly lightens the competition atmosphere a little bit.

9. Combination Missions

It is also my experience that the top teams will combine missions. We complete all the missions in 8 sorties.

10. Attachments

As I've said before, teams need to learn and *think* about how to construct robust structures! Lego doesn't hold itself together under forces in all directions!

Josh

linda zoe said...

My team competed on one pair of tables where the satellite was tilted away quite a lot, making the lever higher off the mat. The robot drivers noticed this during the prep time, and fortunately their plan was to trigger the satellite with a container. They quickly modified their container, taking LEGOs off the sides and building up the height on the front to reach the lever. I was very proud of them for solving the problem on the fly.

Ethan Steckmann said...

1. Table construction
Our competition host built very nice tables but they didnt follow the rules and this killed some teams in the mars mission

3. Sleep Timer

All i can say is OMG that would stink but really they just need to learn the hard way sometimes.

4. Rescue

Team members are allowed to rescue

5. Sensors... again

The built-in motor encoders are great but ya i would love to see next years compatition with more followable lines and stuff that would make a team want to use sensors.

its worth noting that the walls also make great sensors

6. Gears

the drive motors dont need them but some attachment motors do. also gear are inefective and you loss pertion and power

7. Professional behavior

Team seemed more professional and less gracius professional this year then in past years

9. Combination Missions

this is a huge must if you want to get them all dont in under 2:30 but you do loss percision with each mission you add

10. Attachments

atachment we very simple this year we want less missions that are harder

mndstrmmstr1

Joshua Heinzl said...

"its worth noting that the walls also make great sensors "

A wall is not a sensor, but utilizing it is landmark navigation.

"the drive motors dont need them but some attachment motors do. also gear are inefective and you loss pertion and power"

Sounds like you need to have a little lesson on gears. Gears can be used to increase torque or increase speed, but there is a tradeoff.

http://en.wikipedia.org/wiki/Gear

"this is a huge must if you want to get them all dont in under 2:30 but you do loss percision with each mission you add"

If you want a perfect score you have to do them all in under 2:30 ;)

"atachment we very simple this year we want less missions that are harder"

Not sure about other teams but I know our attachments were not 'simple'. They may have had simple concepts but they were not 'simple'.

I'd rather see the same or larger number of missions some/all with increased difficulty :) I'd like to see VERY few to no 400s.

Josh

GeekyTom said...

We suffered from the table height problem as well, but it was a triple failure. Our first mission was combination of tree and wind turbine delivery and tripping the satellite. Unfortunately our trigger arm slipped under the arm on the satellite (something we of course NEVER anticipated) when the robot finally dislodged itself, it took out our trees and turbines at the same time. Luckily my team took it in stride.

Talking Chimp said...

I just have to step in here and say that I find some of the comments by FLL judges here a bit disheartening.

I can look beyond the comments like "you're ignorant or lazy if you don't use sensors," since that is merely a narrow vision of FLL. What upsets me is that some judges are so deeply involved in the tech aspect of FLL that they unwittingly have lost partial sight of the FLL purpose.

I coach a group of regular kids. They're not geniuses, nor are they idiots. However, because it is a school team, we basically took all comers, first come first served. At first, we limited it to 6 kids and "rejected" another 4 who wanted to join. But then we felt that the detriment of expanding to 10 kids was outweighed by the benefit the additional 4 kids would receive from participating. So, we expanded to 10 kids, despite recommendations to the contrary.

One of our kids is ADHD or possibly on the functional autism spectrum. Another is so highly intelligent that he has great difficulty working in groups (or even with just one other kid). Another has a great mind for programming, but cannot assert himself strongly in a group. They are the WORST kids to deal with, and they are my favorite kids to deal with, for just those reasons.

It was not a great experience for us coaches. Very much like herding cats. But every kid on our team had moments here and there when the light would go on, and you could see the moment of discovery in his/her eyes. Those moments were worth all the occasional incidents of having to "time out" them for behavior.

Their approaches to the missions are pedestrian, to say the least. Tools are crude, programming is elementary (push and return), and their overall strategy is fragmented.

Still, we managed to eek out 210 points, which was good for us.(A perfect run would have been 260). The judges at our table were wonderful, as were the ones at the project presentation. From what I gather, the interview went fairly well.

But when I read comments on this blog that criticize teams for not striving harder for excellence or not going beyond just the points, it is disappointing. I understand what you're trying to say, and I agree to a large extent. And I know that in the few minutes that you have with a team, it is impossible for you to know their back-story. '

Still, I would like judges to know that there are many teams like mine, in which the kids are not and probably never will be technically very proficient. But, the little leaps they make from novice to beginner are tremendous, not just in my eyes, but those of their parents, too.

Not all of us are seeking to discover the full potential of the Mindstorms robot. We're just there to have some fun and teach the kids some life lessons.

Having said all that, I did teach them the touch, light, and ultrasound sensors, after reading judges' comments on the lack thereof. But the kids soon learned that the sensors inhibited their progress for the same reasons stated by others in these comments.

They learned how to use the sensors. What's the point of showing it, if it lowers your score? As long as the kids learned it, I don't care if the judge knows it. We're coaching for the kids, not the judges.

Judges: I and the parents appreciate in spades all that you do. You teach our kids, and you teach us coaches/parents. But please remember, when finding fault in a team, that perhaps those faults are compensated by a team dynamic that developed during the 10 weeks of practice. Some teams, like mine, are there for the 10 weeks of practice, and not necessarily the final result. Judges are not expected to know the growth that the team or its individual members experienced. But many times, that is the result of which we are the most proud.

Thanks, judges. You do great work. But please be a little gentler and remember that some of us are just regular folk.

- Coach Dad

EPILOGUE: The judges at our tournament recognized the non-robot qualities of our team. With our paltry 210 (bottom 50th percentile) and our very elementary Tribot and push-return programming, we still received an invitation to the State Championship. Lowest table score of all that advanced. Thank you! Thank you!

Jim Kelly said...

Coach Dad,

I apologize if my postings came off as criticisms of novice teams - that was not my intent. I was just posting observations, take them or leave them, that I thought might benefit some teams or at least just start some discussion.

One of your statements, though, "They learned how to use the sensors. What's the point of showing it, if it lowers your score?" has me in a quaundry, though.

You are correct that using sensors might cause a team to have to change its strategy so much that it suffers points. But I have to ask the question, what is your goal with FLL? Is it to have the students learn some new skills, try new methods, and explore the world of science, math, and engineering or is it just to win a contest?

I know - we all love to win games and come in 1st place. But I think FLL sometimes suffers from the "win win win" mindset and forgets that a large portion of the endeavor is to simply learn and explore.

Your points are well taken and your team deserves a congrats for the scores it received. New or old teams, though, need to focus less on 1st place and more on just the road to getting there... 1st place is just icing on the cake if the team walks away with some new skills and appreciation for real-world engineering.

Jim

Talking Chimp said...

Quandary clarification: What I meant was that the kids did learn how to use the sensors; so, they gained by learning the skill. If one goal of FLL is to "learn and explore," then mission accomplished.

There was nothing to gain by demonstrating at the tournament that they had learned it, other than just to show the judge. The greater value is for the kids to learn, not necessarily to show at the tournament what they have learned (although, that would be nice, too). Like you said, our focus was on the road getting there.

What I'm trying to say is that I care less about what the judge knows our kids can do, than I care what the kids know what they can do.

We did not sacrifice learning for points. We got both. (We did use a touch sensor for the Oil Platform mission.)

linda zoe said...

There has been an interesting idea on the FLL forum recently. To reduce the number of perfect scores, and to encourage use of sensors and more complicated programming, someone proposed a point penalty for every time the robot returns to base. The penalty would be on an increasing scale: 1 point deduction for the first return to base, 2 points for the second return, 3 points for the third, etc. Thus, five returns to base would be 5+4+3+2+1 or 15 points deduction. The individual missons could still be easy enough for rookie teams to do, but the challenge of doing multiple missions in one run would challenge the advanced teams.

I am still mulling this idea over. I think this could be good. Some thought would need to be given to the current "robot rescue" penalty: keep it the same, or drop it and use the "return to base" penalty only.

Ethan Steckmann said...

I dont need to learn about gears i know how to use them and what they do but every gear you use you loss a tiny amount of energy.

So what i said still stands gears loss you ENERGY and gain you power OR speed


Also the walls are landmarks but you can you the nxt motor encoders to sense when you are lined up on the wall.


mindstormMaster1

Joshua Heinzl said...

"So what i said still stands gears loss you ENERGY and gain you power OR speed"

"the drive motors dont need them but some attachment motors do. also gear are inefective and you loss pertion and power"

Energy and power are different ^_^ I believe what you mean is you can gain torque OR speed.

The power that is lost in a geared system is nominal. This is a bad reason for not using gears as they are effective when needed.

"Also the walls are landmarks but you can you the nxt motor encoders to sense when you are lined up on the wall."

Exactly, which means the wall is a landmark and your encoders are sensors.

Josh

Anonymous said...

How about documenting some of those ideas (from #8) you saw that you wouldn't have thought of.and putting them in this blog so everyone can benefit!

Anonymous said...

Recently a poster with the initials FJ has been spamming the blog trying to post an item that was chosen to be deleted.

This person has stated that he/she was a judge for FLL. If anyone knows who this person is, please contact us. It's a shame that an adult has to behave like this - and this is a person who is most likely considered to be a role model by some.

The comment this person is attempting to upload is now being automatically deleted by a Rule but we would sure like to know who is doing it so we can take appropriate action.

The NXT Step Team

Related Posts Plugin for WordPress, Blogger...